It is well known that neuronal network activity is shaped by extrinsic neuromodulation, synaptic interactions and by the intrinsic properties of each neuron within the network. Intrinsic properties in turn are determined chiefly by the ionic currents expressed by each cell. The changes and regulation of each one of these processes results in a diverse repertoire of network outputs. Neurons and networks have been shown to generate stable electric activity despite wide variability in ionic current levels. Such stability could, however, be compromised if variability is allowed to go unchecked. The global ionic current variability in a neuron could be reduced, and output stability enhanced, if the conductance variance of multiple ionic currents depended on each other and were coordinately regulated. Regulation of ionic current levels can in principle be controlled by two classes of mechanisms: 1) mechanisms that sense a departure of activity from a given set point or range that trigger compensatory changes leading to activity restoration, 2) mechanisms that stabilize activity in an activity-independent manner. The crustacean pyloric and gastric mill networks of the stomatogastric ganglion have been used as model systems to study the role of neuromodulation, synaptic properties and intrinsic neuronal properties on the generation of rhythmic activity. These networks generate rhythmic activity patterns that drive digestive behaviors. Other rhythmic pattern generating networks drive behaviors that are also essential for survival (e.g. respiration, locomotion) or are thought to be key in cognitive functions (attention, memory, etc). Because of their basic nature, it could be argued that these rhythms need to be stable and able to recover from disruptive perturbations to maximize survival. The pyloric network has this kind of robust behavior and will be used to examine biophysical mechanisms that stabilize network output. The guiding hypothesis of this proposal is that neuronal and network activities are regulated by two distinct mechanisms at two different time scales: 1) via slow-acting neuromodulatory effects that control the levels and the correlated expression of multiple ionic currents that are not acutely modulated by them, 2) via fast-acting activity-dependent mechanisms that regulate ionic currents levels. I propose to examine the mechanisms of action of these two regulatory processes, characterize their effects in individual neurons, and examine their role on rhythmic activity generation and stability. We will focus especially on the novel, slow, neuromodulator-mediated process. We will use electrophysiological, molecular and computational methods. The capacity to generate stable neuronal output and to recover such output following disease or trauma is crucial to ensure behavioral stability and, ultimately, survival. The mechanisms underlying such stabilization and recovery of function are not well known, and their understanding may be of enormous therapeutical relevance.